Integrating natural language processing in user experiences and interfaces

What I’ve learned from NLP and AI tools that aren’t chatbots

Alicia Drinkwater
Bootcamp
Published in
5 min readJan 2, 2025

--

A woman with projections of code and binary across her face.
By ThisIsEngineering via Pexels

Natural language processing is becoming more and more common in digital experiences, from chatbots to dynamic copywriting. UX has always been concerned with using intentional, if not natural, language and copywriting, but trends are pushing UX design to focus on natural language as a tool in interaction design.

Natural language processing tools can summarize text, search for materials, highlight important words or phrases in text, and extract keywords. When done correctly, these NLP use cases can also make an experience feel more warm and personal.

Here are some ways to use natural language in your UI:

  • Make your experiences more intuitive and easy to understand
  • Make it easier for users to transition between their actions, automation, and AI agents
  • Make it easier for users to learn and remember complex workflows
  • Make your experiences more accessible by removing the need for users to rely on specific UI elements
  • Make your experiences more efficient by inferring data and actions from the user’s context
  • Keep users engaged by saving their progress and allowing them to pick up where they left off
  • Make your experiences more customizable to fit the needs and motivations of each user
An illustration of a man in front of text boxes and lines

NLP in user experiences and interaction design

Most jobs to be done, especially in B2B products, require multiple steps, multiple people, and often multiple days. Sometimes, the user is left waiting for someone or something else, which can lead to a lot of frustration and a risk of forgetting to follow up.

Reminders and notifications in traditional UX can help with this issue, but by using NLP, products can manage a much greater amount of subjectivity and detail. Their ability to store contextual information is part of what makes them so useful: recalling details and proactively introducing new topics makes it feel like you’re chatting with an expert or an assistant.

These interactions don’t have to be limited to chatbots. I personally believe they’re more powerful when used within an interface, blurring the lines between traditional UX and AI-powered experiences. Querying, triggering actions, navigating, and prompting user actions could all benefit from NLP without the need for a full-on chatbot.

An illustration of two people interacting with a larger than life website.

Imagine instead of setting up fine-tuned filters for a home search in Zillow, you could simply describe your home in natural language — a medium-sized Cape Cod with a well-landscaped front yard and an open concept living area — and Zillow, using your prompt, your profile, and your saved homes, would create a custom alert for you. It’s like having a tiny (and limited) real estate agent in your pocket.

The concept of logging in to check on something fades away with an AI agent managing the process. Users wouldn’t need to look things up or search for an item anymore. They’d have to develop a level of trust with their AI agent and rely on them for accurate, timely, and proactive information.

When implementing NLP, there are a few things to keep in mind.

Using natural language in user experiences can create ambiguity if not done carefully. Designers should integrate contextual cues for both the NLP and the user.

User trust is also a challenge in NLP: especially with older generations, computers are often seen as tools that need to be told exactly what to do. Leaving ambiguity to a computer feels unnatural. The UX needs to create a deep level of trust with the user so that they believe the agent can handle the questions they ask. I think this could be achieved through accurate information (proving the AI’s intelligence and competence) and careful questioning and honesty (proving transparency when the AI is unsure, showing that it won’t randomly generate false information).

An illustration of two women sitting on floating message bubbles.

When designing an NLP agent, it’s also important to consider the style of prompts to optimize accuracy. For instance, if the NLP primarily handles search queries, the designer should be aware of the difference between “show me all documents that contain the word ‘architect’” and “documents containing ‘architect.’” The complexity, ambiguity, and structure of the queries matter, although it’s become less important with advancements in NLP technology.

Even though product and tech folks might have been using AI for a while, it’s still emerging in the mainstream. It’s important to remember how little we truly know about the implications of AI within products and to empathize with how unfamiliar many users might feel in an AI-powered experience. To me, these are more reasons to aim for integration and to leverage NLP early in an AI implementation!

As I start implementing some of my first integrated NLP experiences, I’d love to hear from others who have directly integrated AI into a user experience. Were the outcomes what you had hoped for?

References

The Intersection of NLP and UX/UI Design: A Comprehensive Guide

Natural language interface

Natural Language UI — The Next Big Leap in UX?

Using Natural Language Processing to Improve UX

I’m a product leader who loves working with startup teams to help them find their way to product-driven success. I believe in creating sustainable, innovative, and empowering product strategies focused around the humans that build and use products. I also offer coaching and consultation services to build careers, skills, processes, and products. I’ve been involved in dozens of launches and product awards, and I’m always excited to learn and grow with others in startups, product, design, and tech. Feel free to reach out to me through my website or LinkedIn profile. I’d love to chat!

--

--

Bootcamp
Bootcamp

Published in Bootcamp

From idea to product, one lesson at a time. Bootcamp is a collection of resources and opinion pieces about UX, UI, and Product. To submit your story: https://tinyurl.com/bootspub1

Alicia Drinkwater
Alicia Drinkwater

Written by Alicia Drinkwater

Product, design, startups, and strategy

No responses yet